The primary objectives of the proposed study are: 1) development of optimum prognostic scales for use with hospitalized schizophrenic patients; 2) statistical "retrieval" of the Astrup system of diagnostic classification; and 3) determination of the effectiveness of both the Phillips and Becker-Wittman prognostic scales in predicting long-term outcome in the schizophrenias. The above major aims, as well as several minor ones, would be accomplished on the basis of detailed review and ratings of 515 charts of schizophrenic patients hospitalized at the Henry Phipps Psychiatric Clinic between 1948 and 1960. Long-term follow-up data are already available on 350 of these patients, and the patients' hospital charts contain a wealth of background, demographic, and clinical information. Diagnostic classifications by the Norwegian psychiatrist, Christian Astrup, are also available for each patient. To fulfill the above objectives, a variety of multivariate statistical analyses would be employed, including factor analysis, multiple linear regression analysis, and multiple discriminant analysis. It is anticipated that the results will shed light not only on the most effective predictors of outcome in the schizophrenias, but also on current issues regarding the most meaningful system of diagnostic sub-classification to employ. The relative importance of "longitudinal" as opposed to "cross-sectional" symptoms or variables in predicting long-term outcome would also be determined.